Skip to content

Windows: daemon stop times out after 300 s while the warm pool remakes an Android seed (EBUSY on cache.img) #1700

Open

Description

@V3RON

On Windows, simlock daemon stop failed after 300 s while the warm pool was making an Android seed:

PS> simlock daemon stop
{"error":{"code":"SERVICE_FAILED","message":"The running daemon did not stop within 300 s."}}

The daemon log shows the stop asked at 08:26:07 and the daemon stopping only after the seed make failed, about 300 s later:

{"timestamp":1794990367512,"level":"info","module":"daemon.server","message":"Daemon stopping","fields":{"reason":"requested"}}
{"timestamp":1794990669562,"level":"warn","module":"daemon.warm-pool","message":"warm pool creation of a device failed","fields":{"model":"Pixel 9","step":"provision","error":"Error: EBUSY: resource busy or locked, unlink 'D:\\sv\\spec1071\\devices\\android\\sources\\seed_5008dc0e-62bc-41bc-9231-0678b6f62474\\simlock.avd\\cache.img'"}}
{"timestamp":1794990670759,"level":"info","module":"daemon.server","message":"Daemon stopped","fields":{"reason":"requested"}}

A simlock daemon start 35 s later worked.

How it got there: the worker had been started with ANDROID_HOME=D:\Android\Sdk, then restarted with ANDROID_HOME and ANDROID_SDK_ROOT unset and the SDK found at %LOCALAPPDATA%\Android\Sdk (a junction to the same folder). The warm pool then remade the Pixel 9 / API 35 seed, and the stop came during that make. Two things look wrong:

  1. The stop waited for the seed make. It did not give up on the make or stop its emulator.
  2. Removing the old seed's cache.img failed with EBUSY. A process still had the file open.

Seen on Windows 11 23H2 (DesktopWindows), main at 2df19f8, emulator 37.2.13, while specifying #1692. Not checked on macOS or Linux.

Expected

simlock daemon stop returns within its bound while a seed is being made: the make is cancelled and its emulator stopped. A seed whose files are held open is refused or retried, and does not hold up the stop.

Part of #1068.

Written by an agent.

Activity

  1. added
    bug:newReported. An agent may triage it (rule 8); the fast track may set it ready (rule 17).
    os:windowsRuns on, or affects, Windows hosts
    on Nov 18, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    bug:newReported. An agent may triage it (rule 8); the fast track may set it ready (rule 17).os:windowsRuns on, or affects, Windows hostsplatform:androidAndroid emulators or devices

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions